Life After Full-time Work Blog

General principles,  applied to life insurance and annuities

Warning: This is a piece for techies, not for ordinary folks!

Recently I’ve come across a number of pieces, or listened to podcasts, about insurance. It occurred to me that the basic principles involving the need for insurance are relatively simple, though they’re often obscured by a mind-numbing amount of detail that includes aspects of personal taxation in the country of origin of the article or podcast.

I’ve set this out before, in different contexts, so you may recall seeing something like this before. As I looked at what I’ve written, it seems to me that the principles are (a) simple, (b) not sensitive to the time of writing, and (c) applicable anywhere. So I’m reproducing the principles in this blog post, and expanding their application to a topic that seems to have gained relevance recently, which is the need (or not) for income streams guaranteed to last for life.

***

Let’s start with life insurance. Why would you need it? In fact, why would you need any form of insurance at all? (Your home, your car, your life, whatever.) It’s because the future is uncertain, and there are some outcomes that can leave us (or our survivors) financially badly off; and even if we can’t completely avoid those uncertainties, we’d like to do something about the negative financial impact if one of those bad events occurs.

There’s a quick common sense sort of way to think about such events, shown in the table below. This considers future risks in two ways. One way is: how likely is the risk to happen? Is there a high probability that it will occur, or is the probability low? The other way is: how big a financial impact will it have, if it occurs? A high or a low impact?

Risks and financial impacts

Here’s what the table says. If there’s a good chance (a high probability) that the event will happen, build the impact into your budget, whether the impact is big or small. Don’t let the event surprise you. After all, it is likely to happen. In the unlikely event that it doesn’t happen, that’ll be a pleasant financial surprise for you.

How about an event that’s unlikely to happen (low probability), but if it does, the impact will be small? Typically, that’s not worth bothering about. Even if it does happen, it won’t disrupt your financial situation much, because its impact is small.  You may even have a small reserve for these unexpected events – I hope you have your Life Happens fund (or fund for emergencies, as it’s more prosaically called).

Now look at the box that represents something unlikely to happen (low probability), but if it does, it’ll have a big impact. That’s the situation you want to avoid, and that’s when insurance is generally indicated. Think of insurance as a form of pooling your risk with the same risk that others face.

For example, fire insurance for our homes. According to statistics available on the internet, the chance that a person’s home will catch fire in any given year is far less than 1%. But if there is a fire, it can do enormous damage. We don’t keep a personal reserve fund large enough to replace the home – that would be the equivalent of owning two homes! So we pool our risk exposure with others by buying a fire insurance policy. The premium (apart from the insurance company’s loadings) is essentially the product of the probability of occurrence and the likely financial impact. And commonly, the product is a small number. So the premium tends to be small and acceptable, and if we don’t have a claim, that’s just fine. We don’t mind losing the premium.

Now apply that principle to your life. The chance of passing away in any given year is small, while we’re working: it doesn’t even reach 1% until we’re close to retirement. But if we have dependents, the loss of working income if we do pass away can be devastating. So a logical step would be to consider buying term insurance that replaces the lost income, the term ending when work is projected to end. And since the lost income declines by one year’s worth every year, a logical consideration might be what’s called a “decreasing term insurance,” where the amount paid on passing away is very large if that event happens when we’re young (because our dependents lose many years of our income) and the amount declines to zero at the projected retirement age. (Yes, you can and probably should complicate matters by projecting inflation, but that doesn’t change the principle.)

Of course there are many other considerations involving life insurance, which potentially has uses beyond the replacement of working income. But I’ll leave those for the experts to explain, particularly since some considerations depend on laws and taxation, both of which vary from one country to another.

***

Now let’s look at these principles in a different context, the context of unexpectedly long life. Emotionally that’s something we all hope we’ll encounter, particularly if we can stay healthy; but from a purely financial perspective (which is the only perspective I’m considering) it’s a risk. Should you buy insurance against it? Well, let’s look at the principles.

The starting point is obvious, yet (in a way) not at all obvious, because it’s invariably ignored. And it’s that insurance only makes sense when there’s a low chance of a financially negative event happening. Apply that to longevity. It means there’s no point considering longevity protection until the chance of outliving the age selected is low.

What does “low” mean? Anything you like, of course; it’s surely a personal decision. But surely it must mean a lot lower than 50%. In discussing fire insurance for a home, we were talking of a less than 1% chance. The same 1% chance applied to longevity would mean insuring against the chance of living beyond age 100 (or thereabouts – it obviously depends on the table used). (Also, I’m talking about the expected survival years starting from birth. If you start from let’s say, age 65, meaning it’s already a long-lived group we’re considering, the survival age would be even higher than 100.)

Gosh, in turn that implies budgeting for living up to that age – very, very expensive! Yes, but that’s what makes the cost of insurance bearable, because it’s only the last 1% of the survival table that we’re insuring against.

OK, then, let’s consider insuring against the last 10% of the survival table. That would be more expensive, obviously, because we’re hedging an event with a 10% probability. What age would that imply? Roughly age 95. So let’s get this straight: the cost of that insurance is becoming noticeable (never mind the actual calculations and amounts: the concept of insuring against a 10% probability intuitively begins to feel high) and we still need to budget for living to age 95.

OK, then, let’s go to a 25% chance of survival. Obviously this would feel very expensive. But what age would it take us to? Answer: roughly age 90. And it would require budgeting for survival to age 90. Both parts (budgeting and longevity protection) would seem very expensive.

Yet consider what longevity protection is actually available. Insurance policies providing protection against the financial condition of living a long time are called “deferred annuities,” and (in countries where they’re available at all) typically the highest age for commencing the insurance payments is 85. (They have a reason for that, but let’s ignore the reason; it doesn’t change the argument.) Typically there’s more than a 40% chance of surviving to that age from birth. If you’ve already survived to age 65, typically there’s a better than 50-50 chance of surviving to age 85. And that’s for a single person. For a couple, the chance is much higher than 50%: typically the chance that at least one member of a 65-year-old couple survives to age 85 is higher than 75%.

Well, you don’t buy insurance against an event that is either 50% likely (for a single person) or 75% likely (for a couple). That’s ridiculous. The cost would be horrendous. And it is. Remember what I said about sensible insurance, near the start: we don’t mind losing the insurance premium if the event doesn’t occur, because the premium is small.

Well, that’s not the case for longevity insurance with payments commencing at age 85. And that’s why people are reluctant to buy such insurance. Losing the premium by not surviving to age 85 would mean losing a lot of money. And so typically the few people who do buy such insurance specify that, if they pass away before 85, the premium is refunded to the estate. Well, my goodness, what that means is that the true premium (the amount that would be lost if the event, the survival, doesn’t occur) is only the interest that the premium would have earned – which is a small proportion of the premium, and therefore buys relatively little actual insurance – and in turn that makes the insurance seem even more expensive.

To sum up: longevity insurance offering to start payments from age 85 is (typically) better budgeted for rather than bought. Not until (let’s say) age 95 does it really make sense to buy such insurance. And that’s not offered anywhere.

How about an immediate annuity?

Same thing. The odds are strong that you’ll be alive next year. Why insure against survival? Budget for survival. It’s death that you might want insurance against, not survival.

No wonder annuity purchases are relatively few. There’s not really an “annuity puzzle,” which is the technical name for the (apparent) mystery as to why so few people buy immediate annuities. They only make sense if your sole goal is to maximize the guaranteed lifetime income you want, and don’t have a problem losing the purchase price if you are unfortunate enough to die early (as Menahem Yaari  established mathematically in 1965.) No wonder even the relatively few annuities that are purchased need to specify something like a return of the remaining premium if the purchaser dies early (thus noticeably reducing the amount of the annual income payable, relative to a policy without that return feature).

Well, this has turned out to be a diatribe. I didn’t intend it, but I don’t withdraw it.

***

Takeaway

Insurance may be valuable when the chance of an event is low but also its financial impact is high if it occurs. Longevity insurance, as offered these days, doesn’t fit these criteria, because the event being insured against has a relatively high likelihood of occurring.

6 Comments